Exploring the role of anti-solvent effects during washing on active pharmaceutical product purity

Washing is a key step in pharmaceutical isolation to remove unwanted crystallisation solvent, rich in impurities, (mother liquor) from the Active Pharmaceutical Ingredient (API) filter cake. This study looks at strategies for optimal wash solvent selection, minimising dissolution of API product crystals while preventing precipitation of product or impurities. Selection of wash solvent to avoid both these phenomena can be challenging but is essential to maintain yield, purity, and particle characteristics throughout the isolation process.
An anti-solvent screening methodology has been developed to quantitatively evaluate the propensity for precipitation of APIs and their impurities of synthesis during washing. This is illustrated using paracetamol and two typical impurities of synthesis during the washing process. The solubility of paracetamol in different binary wash solutions was measured to provide a basis for wash solvent selection. A map of wash solution composition boundaries for precipitation for the systems investigated was developed to depict where anti-solvent phenomena will take place.
For some crystallisation and wash solvent combinations investigated, as much as 90% of the dissolved paracetamol and over 10% of impurities present in the paracetamol saturated mother liquor was shown to precipitate out. Such levels of uncontrolled crystallisation during washing in a pharmaceutical isolation process can have drastic effect on the final product purity.
Whilst precipitation of both product and impurities from the mother liquor can be avoided by using a solvent in which the API has a solubility similar to that in the mother liquor, for example use of acetonitrile as a wash solvent does not result in any precipitation of the paracetamol API or its impurities. However, the high solubility of paracetamol in acetonitrile, would result in noticeable dissolution of API during washing and would lead to agglomeration during the subsequent drying step. Conversely, use of n-heptane as wash solvent for a paracetamol crystal slurry resulted in the highest amount of precipitation amongst the solvent pairings evaluated. This can be mitigated by designing a multi-stage washing strategy where wash solutions of differing wash solvent concentration are used to minimise step changes in solubility when mother liquor and wash solvent come into contact.